Abstract

Vegetation Community Mapping on the Idaho National Laboratory Site
Track: Environmental Management
Authors: Jeremy Shive, Amy Forman, Ken Aho, Jackie Hafla, Roger Blew, Kurt Edwards

The Idaho National Laboratory Site developed vegetation community maps in 1973, 1978, and 1992, however, no quantitative accuracy assessments were conducted. Our mapping process is being conducted using traditional image interpretation methods performed directly within a Geographic Information System (GIS). We digitized vegetation community boundaries in a GIS which allows image-derived data layers (e.g., NDVI, spatial texture), ancillary GIS datasets (e.g., soils, fires), and other imagery to assist with the delineation of vegetation patch boundaries. Working in a GIS environment provides numerous benefits over traditional photointerpretation mapping methods including vector editing tools, dataset overlay and interpretation, polygon topology, and topology validation. Maintaining the vegetation community map in a geodatabase format also allows for quick updates to polygons following disturbances such as wildfire or facility development. Overall map accuracy was high with 70.7% and 94.2% (Fuzzy Level 4) and we suggest these methods may be useful in other arid landscapes.